Zulip Chat Archive

Stream: Is there code for X?

Topic: Right-invariant Haar measure


Aaron Hill (Aug 02 2025 at 21:49):

Do we have some variant of MeasureTheory.Measure.haar, but constructed to be IsMulRightInvariant instead of IsMulLeftInvariant?

Yaël Dillies (Aug 02 2025 at 22:03):

Can you haar on the opposite group Gᵐᵒᵖ?

Aaron Hill (Aug 02 2025 at 22:07):

I ended up using MeasureTheory.Measure.inv (though I had to do attribute [-simp] MeasureTheory.Measure.inv_eq_selfto unbreak simp)

David Ledvinka (Aug 02 2025 at 22:52):

If I understand correctly that means you are in a setting where your haar measure is inverse invariant so you should be able to prove that its also IsMulRightInvariant. I figured this is in mathlib already but I cant find it. In particular I think there should be a IsMulRightInvariant instance assuming IsMulLeftInvariant and IsInvInvariant.

Kevin Buzzard (Aug 02 2025 at 23:24):

If you're not inverse-invariant then another approach would be to push Haar measure along the inverse isomorphism (which is a homeomorphism).


Last updated: Dec 20 2025 at 21:32 UTC